FWIW, here's a visual of how that works.
The Red Crossbill image below was shot with my Z9 at ISO 3600. I imported it into LR and then PS but didn't do any editing and no noise reduction. Then I made a duplicate image and downsized it in PS to the image dimensions of a D7500 sensor and took the same sized 1200x800 pixel crop from each image. That's effectively what happens when you view the different resolution images at 100% zoom in LR or PS (nothing magical about the 1200x800 size, that's just a convenient size to post here on BCG forums but the key is both crops are the same pixel dimensions and represent a 1:1 pixel view of the corresponding image).
Here's what the final cropped images look like side by side at 1:1 pixel view:
View attachment 73539
The left hand image shows how much closer you zoom in at 100% zoom (1:1 pixel view) with the relatively high resolution Z9 sensor and the right hand image shows the level of zoom with the mid resolution D7500 sensor dimensions. Notice how much more visible the noise is in the left hand image due to viewing a higher res image at 1:1.
Here's the percentage of the each image the crop represents using a 1200x800 pixel crop selection
View attachment 73546
View attachment 73544
Bottom line is that as sensor resolution goes up, for what seems like the same 100% zoom level you're really zooming in tighter on the higher resolution image and defects including visible noise appear to increase even though when processed to the same output file size for printing or web use there really isn't a lot more noise. The takeaway from an image evaluation standpoint is that we either learn to accept that high res images look a bit noisier at 100% zoom level or we learn to evaluate them at slightly less aggressive zoom levels to more accurately represent what we'll see in the final sized image.
The flip side is that this also illustrates why noise appears to rise as we crop deeper into a photo. Of course the cropping alone doesn't change the basic image but when we resize the cropped image to the same final pixel dimensions for printing or posting the noise appears to rise the deeper we crop (and subsequently resize for output) compared to less cropping. For instance if you crop an image from full frame sensor size to say DX crop the noise in the size normalized output image goes up a bit more than a stop similar to if you'd shot the image at a stop higher ISO.